The goal of this proposal is to quantify the role of H3K9me3-based heterochromatin gene silencing in pancreatic beta cell development and to modulate such heterochromatin to obtain more mature beta-like cells from human embryonic stem cells. Type I diabetes (T1D) results from an autoimmune depletion of beta cells within pancreatic islets, resulting in a deficiency in insulin secretion, metabolic imbalances that impair health and lifestyle, and a dependency upon exogenous insulin. Recent studies indicate that T1D also involves beta cell de-differentiation. In 2000, the Edmonton Protocol demonstrated that beta cell transplants from cadavers could elicit exogenous insulin-free survival in severe type I diabetics. Yet success typically requires islets from two cadavers, donors are limiting, and patients can regress to insulin dependence over time. Thus, there is an unmet medical need for donor human beta cells. Extensive research over the past 20 years, including from my laboratory, has defined distinct stages of development of mouse beta cells from the endoderm germ layer and identified numerous signaling effectors. Such information has been used to guide beta-like cell differentiation in vitro from human embryonic stem cells (huESCs). Despite this progress, most laboratories experience difficulty generating robust glucose-responsive insulin secretion (GSIS) in terminal cell products. My lab, for instance, observes robust GSIS in only a limited number of beta-like organoids from huESCs, and generating GSIS-competent cells from induced pluripotent stem cells (iPSCs) is an even greater challenge. The lack of consistency from huESCs and the greater difficulty from iPSCs, which can retain an epigenetic memory of their originating cell type, suggests an epigenetic basis. While H3K9me3-based heterochromatic gene silencing has long been thought to constitutively suppress repeat elements in the genome, my laboratory has recently discovered that such heterochromatin is highly dynamic in liver and pancreatic beta cell development and that such heterochromatin also marks the genes that are the most difficult to activate in various cell reprogramming approaches. Indeed, we present evidence that H3K9me3 heterochromatin is not appropriately modulated in beta-like cells derived from huESCs. We propose two Aims to exploit these insights: Aim 1: Quantify the role of H3K9me3 heterochromatin during beta cell maturation in vivo and in vitro. Aim 2: Employ a knockdown screen of broad and specific heterochromatin modulators during huESC differentiation to beta-like cells to improve glucose-stimulated insulin secretion in vitro and in transplanted animals.